April 9, 2012
ROBERT ANDERSON
Chair, Academic Council
Subject: Report of the Senate-Administration Taskforce on Faculty Salaries
Dear Bob,
On March 12, and April 2, 2012, the Divisional Council (DIVCO) of the Berkeley
Division discussed the report of the Senate-Administration Taskforce on Faculty
Salaries, informed by reports of our divisional committees on Academic Planning
and Resource Allocation (CAPRA), Budget and Interdepartmental Relations (BC),
Faculty Welfare (FWEL), and Status of Women and Ethnic Minorities (SWEM).
DIVCO and the reporting committees all agree that there are serious faculty
salary and compensation issues that must be addressed. UC faculty salaries lag
behind those of our competitors, substantially so in some fields. Additionally,
FWEL noted: “An emphasis on recruitment and retention has led to a serious
loyalty penalty problem and to salary inversion problems in many units.”
We do not, however, believe that this proposal is best strategy for addressing
these issues. We considered the relative benefits and costs of the proposal,
including purported improvements in faculty retention. The prevailing sense of
DIVCO is that the costs of the proposal outweigh the modest benefits we can
expect from its implementation. While there was a broad consensus on DIVCO
opposing the report’s recommendations, we also acknowledge there will
undoubtedly be a subset of our campus colleagues who support them.
The discussion in DIVCO brought a number of serious concerns to the fore.
Berkeley’s approach to compensation
The Berkeley campus currently has a vigorous approach to compensation and
salary setting. Initiatives currently underway include a salary adjustment linked
to promotion to tenure, and a targeted decoupling initiative to provide
competitive salaries to high-performing faculty members. Further, there is strong
Senate involvement from conception through detailed implementation of these
efforts, as opposed to the more mechanistic and centralized approach envisioned
by the proposal.
We believe that the issue on this campus is one of insufficient funding for
competitive salaries, not the process by which salary and compensation is
determined. In addition, the scope and nature of the Senate role in determining
compensation for faculty on our campus is a core principle that must be
preserved going forward.
Merit-based salary setting
We are also concerned that the proposal runs counter to Berkeley’s strong
tradition of merit-based salary setting. DIVCO underscored FWEL’s observation:
Its operation would give faculty extra bumps in salary at the
time of a merit increase essentially based on retention and
recruitment packages offered to colleagues in a broad range of
departments, so that faculty become free-riders on the success of
their colleagues. This is not a merit-based system of adjusting
faculty salaries, and would meet considerable opposition on
many fronts. It has a "Lake Wobegon" aspect to it—that everyone
is above average—which would not sit well with many
observers and which is inconsistent with our traditions.
We believe that any retreat from the merit-based approach that has served this
campus well should be strongly opposed.
Exacerbates recruitment, retention and equity issues
With respect to the proposal’s effect on faculty recruitment and retention,
DIVCO agreed with the BC’s analysis. In sum:
Step 2 of the proposed salary plan significantly raises the cost of
recruitment and retention efforts (assuming no reduction in
quality in the former activity and no “gaming” the step system
for either activity). This indirect cost is substantial—it will likely
raise the costs of recruitment and retention by between 65 and
120%. One possible response to such an increase is that the
campus will forgo recruiting and retaining high-performing
individuals. Another is that it recruits less distinguished
individuals. Yet a third is a distortion (“gaming”) of the rankand-step system to avoid paying that indirect cost, leading to
inequitable assignments of faculty to rank and step. None of
these consequences are desirable and they would adversely
affect Berkeley’s quality, or the morale of its faculty, or both.
While equity was not included in the Taskforce charge, SWEM discussed the
recommendations, noting:
… there is ample preliminary evidence (e.g., the Yahr report;
ongoing salary equity studies in the School of Public Health) that
pay disparities between faculty members at comparable rank
and step correlate with respect to gender …
!
"!
The Taskforce recommends a system by which faculty
advancement to a new rank and/or step is rewarded with a
salary at least equal to the average of all campus faculty at that
rank/step. The effect of such an averaging mechanism for
determining salary would be to reduce disparities between
faculty members; as such, it would necessarily reduce those
disparities that specifically correlate to gender …
Because of their effects on salary inequities, SWEM believes that
the Taskforce recommendations would be positive. We are,
however, adequately worried by the Chair of Budget
Committee’s argument that implementing the Taskforce
recommendations would lead to the discouragement of
recruiting and retaining highly paid faculty. If so, this will also
hurt the recruitment and retention of highly paid women and
URM faculty.
Campus needs beyond salary
DIVCO also noted that the Taskforce recommendations must be considered in
the larger context of campus needs. We believe that we should balance the need
for higher faculty salaries with other pressing campus needs. CAPRA expressed
this well:
We agree with the Task Force that that this under-funding of
faculty salaries has a negative effect on faculty welfare and risks
the loss of faculty to competitors. However, we note that salary
is only one of several factors affecting our ability to compete for
new hires and to retain faculty who have other employment
options. Additional factors that are highly important are the
collegial environment and policies of shared governance for
which Berkeley is rightly famous, and the quality of the
resources available (laboratories, libraries, IT, support services.)
CAPRA has been concerned for some time that we are not
investing enough in these latter resources, and that this
underinvestment has a negative impact on faculty welfare. We
believe that the broader picture should be weighed in evaluating
how to allocate current and/or new funds on the campuses, and
in advocating for increased funding.
UC financial initiatives
Recent UC initiatives such as funding streams and rebenching have been
welcome steps toward efficiency and flexibility for the campuses. These have
been well received by the Berkeley Division because we believe that local control
over finances gives faculty a much larger say in how funds will be used on this
campus. DIVCO views the Taskforce proposal as an unfortunate step back from
this direction by reinforcing that salary budgeting decisions are to be made by
central authority.
!
#!
Many of the key points of the discussion in DIVCO are discussed in detail by BC
in its report. Accordingly, I am appending the BC report in its entirety. We are
happy to provide the detailed analysis contained in the appendices to the BC
report as well, if that would be helpful to Academic Council.
Sincerely,
Bob Jacobsen
Chair, Berkeley Division of the Academic Senate
Professor of Physics
Encl. (1)
Cc:
!
Alexis Bell and Elizabeth Deakin, Co-chairs, Committee on Academic
Planning and Resource Allocation
Benjamin Hermalin, Chair, Committee on Budget and Interdepartmental
Relations
Yale Braunstein and Calvin Moore, Co-chairs, Committee on Faculty
Welfare
Pheng Cheah, Chair, Committee on the Status of Women and Ethnic
Minorities
Aimee Larsen, Manager, Committee on Budget and Interdepartmental
Relations
Diane Sprouse, Senate Analyst, Committee on Academic Planning and
Resource Allocation, and Committee on the Status of Women and Ethnic
Minorities
$!
University of California, Berkeley
COMMITTEE ON BUDGET AND
INTERDEPARTMENTAL RELATIONS
March 27, 2012
BOB JACOBSEN, CHAIR
BERKELEY DIVISION OF THE ACADEMIC SENATE
RE: System-wide Salary Taskforce Recommendations
We write in response to your request for comments on the recommendations of the system-wide
Senate-Administration Taskforce on Faculty Salaries (hereafter the “Taskforce”), set forth in a
February 3, 2012 memorandum to Executive Vice President and Provost Lawrence Pitts
(hereafter the “Taskforce memorandum”). This document and attached appendices detail our
analysis of the Taskforce memorandum.
To summarize our conclusions: we find that implementation of the Taskforce’s proposals would
not be in the interest of the University of California system as whole. Implementation would
definitely not be in this campus’s interest. If it proves politically infeasible to block
implementation completely, then Berkeley should insist the policy be formulated in a way that
allows campuses to opt out of participating; at the very least, campuses must be allowed to opt
out of “Step 2” of the recommendations. Overall, we find that the costs associated with the
Taskforce’s recommendations—financial, political, and in terms of quality—to be greater than
the Taskforce memorandum suggests; that the costs outweigh the benefits; and that the benefits
are less than the Taskforce memorandum suggests. Moreover, many of the purported benefits
could be realized via less costly means—means that have already proved successful on this
campus. We detail the rationales for these conclusions below. Given the length of this
document, we have divided it into sections and provided short summary paragraphs at the end of
most sections.
Contents
!"#$%&'(&)$*#+,&--.......................................................................................................................... /
0(12$3+-4- ........................................................................................................................................ /
#$%&'(")*+(+",-."/0-.$-1")*-'&%-+ 22222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222 3
4-.$%&'(")*+(+"5%$+$-1"6%*7"8&'%0$(7&-(",-."8&(&-($*-2222222222222222222222222222222222222222222222222222222222222222222222 9
:*;$($',;",-."<(=&%")*+(+2222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222 >
5+6&-4$7&8&9(4-............................................................................................................................... :
4-'%&,+&."/,'0;(?"@,;,%$&+22222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222 A
47B%*C&."/,'0;(?"8&(&-($*-"D5;;&1&."E&-&6$(F 222222222222222222222222222222222222222222222222222222222222222222222222222222222222 G
5"H&,-+"DI0&+($*-,J;&F"(*"K&&B"L)"/,'0;(?"@,;,%$&+")*7B&($($C& 22222222222222222222222222222222222222222222222222222 G
5"H&,-+"(*"L-.*"(=&"M*?,;(?":&-,;(? 2222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222 N
5"H&,-+"(*"47B%*C&"OP0$(? 22222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222222 N
;(8<=$%&><#?-................................................................................................................................. @@
!"
Our Review Process
"
In addition to considering the Taskforce memorandum itself, other information was gathered. To
wit, the Budget Committee (BC) Chair had meetings with (i) Vice Provost Janet Broughton and
EVC&P George Breslauer; (ii) Bob Jacobsen and Christina Maslach, Chair and Vice-Chair of
the Berkeley Division of the Academic Senate; and (iii) Robert Anderson, Chair of the Academic
Senate (system-wide). The BC Chair also attended the March 8, 2012 meeting of the University
Committee on Appointments and Promotions (UCAP) in Oakland, at which the Taskforce
memorandum was discussed. Oral summaries of the discussions at those meetings were
provided to the other members of the BC and discussed by the BC as a whole. We have also
availed ourselves of data supplied by Vice Provost Broughton on retention and salaries.
As part of our review process, we provided a written preliminary assessment to Division Chair
Jacobsen (see Appendix A). This assessment was discussed at the March 12, 2012 DIVCO
meeting. Our preliminary assessment was—as our final assessment is—negative. In our earlier
assessment, we indicated that the recommended policy would not adequately achieve its stated
goals; could prove unduly costly to the campuses; and would very likely significantly increase
Berkeley’s costs of attracting and retaining the best scholars. Further review has not changed
those conclusions.
At the March 12th DIVCO meeting, a concern was expressed that our preliminary assessment
neither discussed nor gave appropriate weight to the positive consequences of the Taskforce’s
recommendations in terms of improved equity. We address that concern below. Although the
Taskforce’s recommendations may improve equity, the improvement is more modest than it
might at first seem.
High Costs
Direct Costs and Funding Concerns
"
In the first year, the Taskforce estimates that annual salary costs would increase by
approximately $30 million system-wide above what they otherwise would have been. After two
years, the cost rises to approximately $60 million. Projections beyond that are not offered, but
since year three would represent the third year of a normal merit cycle, extrapolation suggests
additional annual salary costs would be approximately $90 million. Costs will likely rise
thereafter. Note that these cost estimates exclude the additional costs that could arise as a
consequence of recruitment and retention (discussed below, see also Appendix B).
In discussions, proponents of the new salary system have claimed it will be put in place only if
adequate state funding is provided. Yet some phrases in the Taskforce memorandum create
doubt about whether that is truly a condition for implementation. For example, point #3 of the
memorandum’s executive summary suggests that adequate state funding may not be a necessary
condition; and p. 11 of the memorandum it expressly states that “individual campuses will have
3"
to generate the salary increase dollars.” Moreover, this last excerpt refers to Step 2, which for
the Berkeley campus will be the more expensive of the two steps.
Even if state funding were provided for the first few years of the program, experience teaches
that the state is an unreliable partner. There is a real danger that state funding would later
disappear, and the new salary program would become either an unfunded mandate (even if it
doesn’t start in whole or in part that way) or it would have to be suddenly abandoned, creating
serious equity and morale issues because some faculty would have benefitted by virtue of the
timing of their merit cycles and others would be left with nothing.
Cost estimates for the new salary system are predicated on absorption of decoupled increments at
time of next merit review (see footnote 7 of the Taskforce memorandum).* Under current
Berkeley practice, decoupled increments are not absorbed at time of next merit; rather, actual
salary equals the rank-and-step salary at the new step plus the previously earned decoupled
increment. Absorption means that the decoupled increment is eliminated if the rank-and-step
component at the new step exceeds the faculty member’s current salary; or it is reduced—and
she receives no pay increase—if her current salary exceeds the rank-and-step component at her
new step.† For over a decade, decoupled increments have not been absorbed on this campus
because absorption was deemed to be self-defeating in retention cases (faculty knew that the
increased decoupling due to retention would shortly be “taken back,” so they discounted
Berkeley’s retention offers) and demoralizing when it led to no salary increase despite a
successful merit. If Berkeley were compelled to adopt the Taskforce’s new salary program, it
would either have to resort to absorption, which would reintroduce the aforementioned problems
and would not be popular, or this campus would face even higher costs than currently estimated.
Section summary: The costs of implementing the Taskforce’s new salary program are very
high, and they are likely significantly higher than the Taskforce memorandum indicates. Unless
Berkeley wishes to drop its policy of not absorbing decoupled increments at time of merit
advance, the costs will be higher still. There are reasons to be suspicious of claims that the new
program will be implemented only if adequate state funding is available. Indeed, there is a
significant chance that the campuses will have to come up with the necessary funding.
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
R
"The Taskforce writes about the “off-scale” portions of faculty salaries, which are called decoupled increments on
this campus. Most faculty salaries at Berkeley have a rank-and-step component, which is the amount listed in the
relevant pay scale for a faculty member’s rank and step, and a decoupled increment; the sum of the two equals a
faculty member’s salary."
S
"For example, consider two physicists, Mary and Jane. Both are currently Professor, Step IV. Mary’s salary is
$105,000 and Jane’s $108,000 (their decoupled increments are $5700 and $8700, respectively). Absent absorption,
a one-step merit advance for each would make their salaries $112,100 and $115,100, respectively. With absorption,
they would be $106,400 and $108,000, respectively."
Q"
Indirect Costs Arising from Recruitment and Retention
"
Recruitment or retention cases often award the faculty candidate a larger-than-average salary for
his or her rank and step. Consequently, the recalculated average salary for that rank and step will
be higher. Because, under Step 2 of the proposed salary program, that recalculated average is the
minimum salary of anyone advancing to that rank and step in the future, there is an indirect cost
of recruitment and retention due to those actions’ effects on the salaries of others. That is, if a
faculty member gets ! more in salary than is average for her rank and step, then the pay level for
that rank and step increases by !/N, where N is the number of faculty currently at that rank and
step.‡ This increase in the pay level is an additional indirect cost of the proposed salary program.
An analysis that the BC Chair conducted—subsequently reviewed and endorsed by the
Committee as a whole—indicates that this indirect cost raises the total cost of a recruitment or
retention action by between 65 and 120% of the direct cost. That is, under the proposed salary
program, the cost of recruitment and retention would roughly double. A copy of that analysis is
attached as Appendix B. Some members of the Committee believe those estimates of the
indirect costs are too low because these estimates exclude certain factors (see footnote 2 on page
2 of the Appendix B document).
To provide some context on those estimates, data from academic years 1998-2006 indicate that
Berkeley had approximately 46 retention cases per year and retained on average 33.§ Suppose
that, on average, each retained individual’s salary is $20,000 above what it would otherwise have
been;** that is, above the relevant rank and step average. Hence, the direct annual cost of
retention is $660,000. Even just the 65% estimate implies, therefore, an additional indirect cost
$429,000, roughly the equivalent of four new assistant professors (including benefits). Hence,
in terms of forgone slots, the lowest estimate of the indirect cost arising from the proposed salary
program is equivalent to a loss of 6% of current annual recruitment on this campus.
The Committee believes that such a significant increase in the cost of recruitment and retention
will either discourage recruitment and retention or cause other distortions. A well-known
principle is that if the cost of an activity increases, less of it is done. Hence, a response to the
proposed salary program could be less retention of key faculty or less effort to attract top faculty
to Berkeley. A related distortion could be recruitment of less distinguished faculty, for whom
salaries will tend to be lower.
Yet another possible distortion is that, although Berkeley continues to retain and recruit the same
caliber faculty as before, it seeks to avoid the indirect cost by advancing or appointing these
faculty at very high ranks and steps. The rationale for such a distortion is that the indirect cost
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
T
"Actually, the increase in salary is !/N only for one year. It is higher for subsequent years because the first
beneficiaries of this higher salary also, then, contribute to a higher average. See the analysis in Appendix B."
U
"Data supplied by Vice Provost Broughton. More recent data suggest the number of retention cases is closer to 50
per year and that we retain 38 to 42 per year."
RR
"The figure of $20,000 is a low-end ballpark estimate derived from recent retention cases."
9"
disappears if a candidate is advanced or appointed to a rank and step at which the average salary
is approximately equal to the salary necessary to retain or recruit him or her. Such a
manipulation would make a mockery of the notion that rank and step reflect academic merit. It
would also be demoralizing to long-serving Berkeley faculty, who would risk seeing less
accomplished faculty advanced or appointed at ranks and steps well above their own. Certainly
many long-serving Berkeley faculty would see that as inequitable.
Section summary: Step 2 of the proposed salary plan significantly raises the cost of recruitment
and retention efforts (assuming no reduction in quality in the former activity and no “gaming”
the step system for either activity). This indirect cost is substantial—it will likely raise the costs
of recruitment and retention by between 65 and 120%. One possible response to such an
increase is that the campus will forgo recruiting and retaining high-performing individuals.
Another is that it recruits less distinguished individuals. Yet a third is a distortion (“gaming”) of
the rank-and-step system to avoid paying that indirect cost, leading to inequitable assignments of
faculty to rank and step. None of these consequences are desirable and they would adversely
affect Berkeley’s quality, or the morale of its faculty, or both.
Political and Other Costs
"
Even if adequate state funding were available for the new salary program, it might prove
politically costly (i.e., a public relations disaster that would have negative political
repercussions) to direct so much of that funding to faculty salaries at time when the University is
otherwise making cutbacks in staff and programs and raising tuition and fees.
Such political fallout could be exacerbated by the following. Most of the dollars spent on salary
increases under the Taskforce’s proposal will go to faculty with below average salaries. Because
faculty salaries are positively correlated—albeit imperfectly—with accomplishment, the
Taskforce’s proposal will—on average—more heavily reward less accomplished faculty than it
will more accomplished faculty. Given strong public sentiment for more merit-based pay for
public employees (e.g., school teachers), this aspect of the Taskforce’s proposal could make it
even more politically costly.††
Political concerns aside, there are many other pressing needs on this campus. Hence, although
an increase in faculty salaries is to be welcomed, there are other campus priorities that would
also be very deserving of a share of any increased state funding.
If no or inadequate state funding is provided for the Taskforce’s proposal, but it is nevertheless
enacted, then the consequences could be even more painful, because funds currently allocated to
staff, faculty hiring, infrastructure and maintenance, and ensuring Berkeley remains accessible
and affordable would need to be diverted to fund the proposed salary plan.
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
SS
"In this regard, it should be remembered that one motive President Yudof had to make the October 1, 2011 salary
adjustments partially merit based was precisely to make it more politically palatable than a simple across-the-board
salary increase."
>"
Furthermore, to the extent that the Berkeley campus will need to pay for the new salary program,
it will likely be forced to abandon other salary programs, such as market-based salary
adjustments at time of tenure and the current targeted decoupling initiative (TDI). Faculty who
were promised TDI awards or who were led to expect a significant salary adjustment at time of
tenure are likely to feel they had been treated unfairly. Such demoralization could prove costly
in terms of commitment of effort and future retention cases.
Section summary: There are significant public relations costs to the new salary program even if
adequate state funding is available. Moreover, if such funding is truly available, there are other
pressing needs that should also be addressed. If no or inadequate state funding is provided, then
the new salary program will have a significant effect on other spending on the campus, including
facilities, faculty recruitment, and existing salary programs. These would all prove problematic
and lead to lower faculty morale.
Modest Benefits
Increased Faculty Salaries
"
There is no denying that the new salary program will raise some faculty salaries. Three points,
though, should be kept in mind: (i) the increases are relatively modest on this campus; (ii) the
new salary program is not the only means available for increasing salaries; and (iii) the increases
will not be evenly distributed among the faculty.
According to the Taskforce memorandum, the average salary increase for Berkeley faculty is
2.4% in the first year and 2.7% in the second year. Both numbers, it should be noted, are less
than the 3% we received on October 1, 2011.
The October 1, 2011 salary adjustment or traditional range adjustments (increases to the on-scale
portion of salaries) represent alternative models of providing salary increases. One could also
consider variations on such models: for instance, a more progressive version of the October 1,
2011 salary adjustment that gave an x% increase on the first $y thousand of salary, a z% increase
on the next $w thousand, and so forth, where z < x (e.g., 4% on the first $100,000, 2% on the
next $100,000, and 0% thereafter). Another alternative is to give a percentage increase to the onscale portion of salary (i.e., like a range adjustment) and a smaller percentage (or no) increase to
the decoupled increment.
Under the Taskforce’s proposed salary plan, faculty whose salaries are above average for their
rank and step could see little or no increase at all: as envisioned in the Taskforce memorandum,
any increase in the scale would result in the absorption of an equal amount of those faculty
members’ decoupled increments. For example, consider a Professor, Step IV on the regular
scale who has a decoupled increment of $24,000 (the Berkeley average). Her current salary is,
thus, $123,300. She would be eligible for a salary increase at her next merit advance. The onscale salary at Professor, Step V, under the new salary program is estimated to be $124,100. Her
increase in salary would, therefore, be $800 or 0.6%. Had her decoupled increment been
$25,000, she would have received no salary increase.
A"
Although the highly progressive nature of the Taskforce’s proposed salary program could be
welcomed by some, many other faculty would perceive it as being unfair that they received little
or no increase under the plan, especially after they have enjoyed a successful merit.
Section summary: The Taskforce’s proposed salary plan would raise salaries on average. That
noted, similar size increases could be achieved via alternative programs that would not generate
the indirect costs and other distortions associated with the Taskforce’s plan. Moreover, it is
likely that many faculty will view the Taskforce’s plan as unfair insofar as the distribution of
salary increases is highly uneven and many meritorious faculty would enjoy little or no pay
increase.
Improved Faculty Retention (Alleged Benefit)
"
It is claimed in the Taskforce memorandum (p. 13) that the new salary plan (specifically Step 2)
“is a mechanism that has been in place at UC Irvine for several years, … and has proved
effective in faculty retention.” A priori, we do not understand why this would be true even if it
were. Worse, we have been told that UC Irvine actually suffers from retention problems.‡‡ It is
unfortunate—and rather unscientific—that the Taskforce did not compare retention rates at UCI
with those at the other UC campuses.
As noted above, because Step 2 of the Taskforce’s salary plan significantly increases the costs of
retention, it should lead to fewer successful retention cases. For that reason it could lead to more
successful poaching of Berkeley faculty by our rivals: given the costs of generating outside
offers, rival institutions are more likely to hunt, all else equal, where they think their chances of
success will be greatest. In short, adoption of the Taskforce’s plan could lead to Berkeley losing
more of its stars.
Section summary: Although the Taskforce memorandum claims a benefit of the new pay plan
would be to improve faculty retention issues, both logic and actual experience indicate that just
the opposite is more likely true.
A Means (Questionable) to Keep UC Faculty Salaries Competitive
"
One of the principal goals of the new salary program is to keep UC faculty salaries competitive
with those at relevant peer institutions. The way this would work is through Step 2 of the plan:
Every competitive recruitment and retention case would raise the average salary at the
candidate’s rank and step. Specifically, if she receives ! more in salary than is average for her
rank and step, then the pay level for that rank and step increases by !/N, N the number of faculty
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
TT
"Conversations that Division Chair Jacobsen had with UCI colleagues indicate that UCI has a retention problem in
the sciences. In a conversation with BC Chair Hermalin, Senate Chair (systemwide) Anderson confirmed that
retention was an issue across the campus at Irvine. "
G"
currently at that rank and step. Faculty subsequently advanced to that rank and step would
obtain a greater salary (see page 3 of Appendix B for details). In this way, UC salaries would
increase somewhat to reflect market.
Although this effect will tend to raise UC salaries, observe that division by N means the effect is
somewhat limited if there are a large number of faculty at the candidate’s rank and step. (For
example, if N = 30, then the long-run effect of a ! of $20,000 is a $1,000 salary increase for each
individual advancing to that rank and step.) Of course, if there are lots of competitive
recruitments and retentions at a particular rank and step, then the increase in the salaries of others
advancing to that rank and step will be larger.
A serious concern about this mechanism is that it is likely self-defeating. As discussed above,
because this mechanism roughly doubles the cost of recruitment and retention, it will in all
likelihood lead to either less recruitment and retention, or to a gaming of the system by which
recruits and retained faculty are brought in at or advanced to higher ranks and steps to keep their
high salaries from having much of an effect on the relevant averages. In short, a mechanism that
seeks to operate under the principle that a rising tide raises all boats simultaneously provides
incentives to build a dam to keep the tide from coming in. Moreover, to the extent that cost
concerns cause a campus to lose retention cases involving its high fliers—people whose salaries
are already above average—there is a danger that the mechanism can actually serve to drive
down salaries as water flows out of the harbor.
As with the issue of generally increasing salaries, there are other mechanisms available;
moreover, these mechanisms are more direct and do not have the negative features of the
Taskforce’s recommended plan. Two such mechanisms are the use of targeted decoupling
initiatives (TDIs) and efforts to provide discipline-specific, market-based salary adjustments at
the time of promotion.§§
Section summary: Although it is advertised as a way to keep UC faculty salaries competitive,
the Taskforce’s recommended salary plan is a very indirect way of doing so. Its ability to do
so—even assuming no negative feedback loop in terms of reduced recruitment or retention or
gaming of the step system—seems modest. Once feedback is taken into account, the prospects
for the Taskforce’s plan to enhance the competitiveness of UC salaries become even dimmer.
Furthermore, we reiterate that the Taskforce’s plan is not the only possible answer: in particular,
use of TDIs and market-based adjustments at time of promotion would serve the same purpose,
more directly and with fewer side effects."
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
UU
"Currently, Berkeley has just one promotion market-based salary adjustment, done at the time of promotion to
Associate Professor (tenure). The administration and Budget Committee have expressed an interest in having a
similar program at time of promotion to Professor, should funds become available. "
V"
A Means to Undo the Loyalty Penalty
"
Another claimed benefit of the Taskforce’s recommended salary plan is that it will address the
“loyalty penalty,” (see page 6 of the Taskforce’s memorandum); that is, the tendency of longserving faculty at UC campuses to get lower salaries than newly hired faculty of comparable, or
even lesser, achievement.
To the extent long-serving faculty have lower salaries than more recently hired faculty at the
same rank and step, then they will indeed see an increase in their salaries in the short run under
the Taskforce’s plan. Note, however, that the plan will do nothing to address salary inversion by
age or step. That is, if the new, highly paid faculty are at a lower rank and step, their high
salaries will have no effect on the salaries of long-serving faculty.
The long-run benefits of the proposed salary plan vis-à-vis the loyalty penalty are questionable.
Data made available to us by Vice Provost Broughton show that the probability of becoming a
retention case drops steadily with age after 35.*** If Berkeley’s own hiring shows a similar age
pattern, then recruitment and retention at the upper levels of the professoriat will be relatively
rare and, thus, the influence of those actions on salaries at the upper steps of the professoriat
minimal.
It is sometimes claimed that women suffer disproportionately from the loyalty penalty. The
limited evidence available to us calls that claim into question, at least for the Berkeley campus:
for the period 1998-2006, women faculty were 30% of all retention cases, while they were less
than 25% of the total faculty (source of data: Vice Provost Broughton). Under-represented
minorities were also disproportionately retention cases: for the same time period, nearly 10% of
all retention cases involved members of under-represented minorities, while this segment made
up little more than 6% of the faculty.
Section summary: The Taskforce’s recommended salary plan could partially alleviate the
loyalty penalty in the short run insofar as long-serving Berkeley faculty without recent retention
actions (who are, thus, likely to have below average salaries for their rank and step) will see
immediate raises. The Taskforce’s plan is not, however, a long-run solution to the loyalty
penalty. In neither the short nor long run is it a solution to issues of salary inversion by age and
step. Finally, we again note that it is not the only possible remedy: TDI programs can and have
sought to address the loyalty-penalty issue directly, again without the side effects of the
Taskforce’s plan.
"
A Means to Improve Equity
"
Some of the issues considered previously already touch on matters of equity. One issue
connected to equity is the salary dispersion within rank and step; specifically, if rank and step are
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
RRR
"The data concern Berkeley retention cases from academic years 1998-2006."
N"
supposed to reflect achievement, then equity—at least by some definitions—suggests faculty at
the same rank and step should earn similar salaries. Because the Taskforce’s recommended
salary plan sets a floor for each faculty member’s salary at time of advancement or appointment
to a given rank and step equal to the average at that rank and step, it must necessarily reduce
salary dispersion within each rank and step. Consequently, it will improve equity.
Two questions arise: (i) how inequitable are current Berkeley salaries? And (ii) by how much
would the Taskforce’s salary plan improve equity? In other words, how big is the problem and
how much of a fix will the Taskforce’s salary plan provide?
A complete study of equity was beyond the capacity of this Committee. However, a limited
exercise, while not definitive, does offer some sense of what the answers to those questions
might be. The details of our analysis may be found in Appendix C. For data, we took all
Professors in the Divisions of Arts and Humanities and Social Sciences (with the exception of
Economics, which has a different pay scale) at Steps III through IV.5. The rationale for
choosing this group is given in Appendix C. For each step, various measures of inequality were
constructed using actual salaries and also under the assumption that those with salaries below the
average would have them raised to the average. Three measures were calculated: the coefficient
of variation (CV), the inter-quartile range divided by the median (a “robust” variant of CV—call
it RCV), and a Gini coefficient. For all these measures, a higher score means greater inequality.
Across the four steps, using actual data, the CV varied between 10% and 24%; RCV between 9%
and 24%; and the Gini coefficient between .06 and .12. To provide a sense of magnitudes, the
OECD iLibrary reports a CV for US income of 90% and a Gini coefficient of .38. For Sweden,
these numbers are 80% and .23. Across the four steps, raising all salaries below average to the
average, the CV varied between 6% and 16%; RCV between 1% and 8%; and the Gini
coefficient between .03 and .07.
It is doubtless open to debate as to what constitutes an unacceptable level of inequality within
rank and step and how valuable the reductions in that inequality described above are. That said,
given the various competing goals that the campus faces—including preserving and enhancing
the quality of its faculty—we are unconvinced that these equity gains trump the significant costs
that the Taskforce’s proposal would impose.
Finally, we again note that the proposed pay system is not the only way to address inequality:
TDIs are another proven way to address it.
Section summary: The Taskforce’s proposed salary plan would reduce inequity in salaries with
any given rank and step. Although a legitimate concern, a partial analysis of the situation at
Berkeley suggests that such salary inequality is not especially great. Moreover, the Taskforce’s
plan does not seem as if it would reduce that inequality dramatically. Finally, alternative
methods exist to address inequality, such as the use of TDIs.
!W"
Final Remarks
"
After careful review and discussion, the Committee has concluded that the new salary program
set forth in the Taskforce memorandum is a bad plan. It is very costly and the costs—claims of
its proponents not withstanding—will be borne in large part, if not entirely, by the campuses. In
contrast, the benefits it delivers, at least for the Berkeley campus, are relatively small and no
greater than those provided by existing, less expensive programs. Whatever new funding may
exist to fund the Taskforce’s plan could be better spent—if only to expand existing salary
programs such as market-based salary adjustments at time of promotion and targeted decoupling
initiatives (TDIs).
It is not only that the costs of the proposed plan would outweigh its benefits that make us
recommend against it. The proposed plan would also generate a number of distortions that
would jeopardize the quality of UC Berkeley, demoralize its faculty, or both. It would—on
average—give greater pay increases to our less-accomplished faculty than to our moreaccomplished faculty, with the latter group possibly facing small to no pay increases in the short
term. It would likely mean the end of existing salary programs, which would represent broken
promises. Finally, it would mean either Berkeley would be seriously disadvantaged in terms of
recruitment and retention because the costs of those activities would roughly double, or, as a
means of evading those costs, Berkeley would be forced to “game” the rank and step system so
that rank and step came to reflect market salary more than merit.
For all these reasons, our recommendation is that the Senate resist as fully as possible the
adoption of the Taskforce’s recommended salary plan. Should that prove a losing battle, then the
line of retreat should be to allowing campuses to opt out of it. The final line should be that
Berkeley be allowed to opt out of the Step 2 provision, which is the most pernicious and harmful
feature of the plan.
Benjamin E. Hermalin, Chair
Committee on Budget and Interdepartmental
Relations
BEH/mg
!!"